AI Transformation: what the research shows
AI reaches a local government as a technical change, a governance question and a human one at the same time. Each of the three figures below opens one side.
What it means for the people a leader leads
For staff, AI can arrive as a new tool, an open question about the rules and a public concern at the same time.
Homepage signals
Suspected AI-assisted records requests
Emailed AI drafts awaiting review
Who can access camera data?
Situations
A records request appears to use AI. Staff need to clarify which records are sought.
An AI-produced draft arrives by email. The reviewer needs to establish what was checked and who owns the recommendation.
Police license-plate cameras raise questions about who can search, what is retained and how use is reviewed.
Evidence. A 2026 Maine municipal survey reports requests described as AI-generated. A county policy assigns human verification of AI-produced work. Two cities document license-plate camera oversight.
Limits. These sources establish reported experience and governance responsibilities, not national frequency. Camera oversight is a separate technology question; not every camera uses AI.
Sources
How to read the three figures
Each side has its own evidence. The three figures above come from two publishers and two groups, U.S. employees and U.S. adults, so they sit side by side and are not ranked against each other.
The technical side: how many employees have used AI at work
It was 21% in the second quarter of 2023 and 40% in 2025: seven readings over three years.
The figures behind the chart
Percent of U.S. employees who have used AI at work, Gallup
| Period | All employees, any use | Frequent use | Daily |
|---|---|---|---|
| second quarter of 2023 | 21% | 11% | 4% |
| second quarter of 2024 | 27% | 12% | 4% |
| second quarter of 2025 | 40% | 19% | 8% |
| third quarter of 2025 | 45% | 23% | 10% |
| fourth quarter of 2025 | 46% | 26% | 12% |
| first quarter of 2026 | 50% | 28% | 13% |
| second quarter of 2026 | 52% | 30% | 15% |
For the people a leader leads
Staff are among the U.S. employees this figure counts. For many of them, AI is already a tool they have used at work.
More readings: frequent use and businesses
In the second quarter of 2026, 52% of U.S. employees had used AI at work at least a few times a year, 30% a few times a week or more and 15% daily. Three years earlier the three figures were 21%, 11% and 4% (Gallup).
Businesses are a different population from employees, asked a different question. In the two weeks to 3 May 2026, 19.8% of U.S. businesses said they were using AI in any business function, and the rate ran between 17% and 20% from December 2025 to May 2026; among firms of 250 or more it was 37%. The question changed wording in November 2025, so no earlier figure joins it (U.S. Census Bureau, Business Trends and Outlook Survey). No comparable local government series exists. An employee using a tool and an organization adopting one are different measures, and no survey here compares the two.
The governance side: the plan employees hear about
It was 15% in May 2024, 22% in May 2025 and 24% in August 2025. A communicated plan is what employees report hearing. It is not a measure of whether a policy exists.
The figures behind the chart
Percent of U.S. employees, Gallup
| Period | Clear plan communicated |
|---|---|
| May 2024 | 15% |
| May 2025 | 22% |
| August 2025 | 24% |
| May 2026 | 25% |
For the people a leader leads
Where staff have not heard a plan for AI, the rules can end up settled desk by desk.
No national series measured AI policy inside local government with a reading in 2026 when this page was written. Governance is the part of AI a council can see and ask about: what the organization uses, who is accountable, what residents’ data it touches and what it will not do.
The human side: the room where it gets discussed
It was 37% in 2021 and 38% in 2022, then 52% in 2023 and 50% to 52% in every year since. 37% were equally concerned and excited in 2026, 9% more excited.
The figures behind the chart
Percent of U.S. adults, Pew Research Center
| Year | More concerned | Equally | More excited |
|---|---|---|---|
| 2021 | 37% | 45% | 18% |
| 2022 | 38% | 46% | 15% |
| 2023 | 52% | 36% | 10% |
| 2024 | 51% | 38% | 11% |
| 2025 | 50% | 38% | 10% |
| 2026 | 52% | 37% | 9% |
For the people a leader leads
Staff are among the U.S. adults this figure describes. The concern, and the excitement, may already be in the room before AI is on the agenda.
Whether a concern about AI can be said out loud inside an organization is a Face the Truth question.
A second measure: whether AI does more harm than good
of U.S. adults said in May 2026 that AI does more harm than good
The share was 40% in 2023, 31% in 2024 and 31% in 2025. In May 2026 the most common answer, given by 52% of U.S. adults, was that AI does about equal harm and good.
The figures behind the chart
Percent of U.S. adults saying AI does more harm than good, Gallup and Bentley University
| Year | More harm than good |
|---|---|
| 2023 | 40% |
| 2024 | 31% |
| 2025 | 31% |
| 2026 | 39% |
Three questions only the organization can answer
No national series measures any one city or county. These three questions are for a leadership team to answer about its own.
- Which AI tools are staff using today, for what work, and which of those uses does the organization’s guidance cover?
- Who owns the AI policy, and when did staff last hear what it says?
- Where can someone raise a concern about AI without it sounding like resistance, and who opens that conversation?
Further reading
- AI use, June 22, 2026: What Stalls AI Adoption in Local Government?
- AI and privacy, March 5, 2026: What a Federal AI Privacy Ruling Means for Local Government
- AI and judgment, January 16, 2026: The Question AI Can’t Answer
Primary sources
Data 2021 to 2026. Gallup, Pew Research Center, Gallup and Bentley University, and the U.S. Census Bureau.
- Gallup: Employee Engagement Remains Flat as AI Adoption Accelerates (July 2026)
- Gallup: Artificial Intelligence Indicator
- Pew Research Center: Young Adults in the U.S. Are Increasingly Wary of AI (August 2026)
- Gallup and Bentley University: Americans Cool Toward AI (2026)
- U.S. Census Bureau: AI Use by Businesses, Business Trends and Outlook Survey (May 2026)